VXLAN MP-BGP EVPN L2 VNI

In the first two VXLAN lessons, we used static ingress replication and multicast to learn MAC addresses. These are two flood-and-learn options that operate on the data plane. In this lesson, we’ll examine how to use MP-BGP EVPN to learn MAC addresses for VXLAN.

VXLAN MP-BGP EVPN separates the control plane (MP-BGP for MAC/IP advertisement) from the data plane (VXLAN encapsulation). We use MP-BGP to advertise MAC and IP address information between leaf switches. This eliminates the need for flood-and-learn mechanisms for unicast traffic and is a more efficient and scalable option.

EVPN uses a route distinguisher and route-target, similar to MPLS VPN. The RD is what makes a route unique. The RT defines where we want to export and import these routes. On Cisco NX-OS, the RD and RT are automatically generated by default.

Despite MP-BGP EVPN’s advantages, we still need something for Broadcast, Unknown Unicast, and Multicast (BUM) traffic. You’ll need to configure static ingress replication or multicast for BUM traffic. In this lesson, I’ll explain how to configure MP-BGP EVPN, and we’ll use multicast for BUM traffic.

Configuration

Here is the topology I’ll use:

Vxlan Spine Two Leafs Cisco Nx Os

I’m using Cisco Nexus 9300v switches running version 9.3(9). The hosts are simple Ubuntu Docker containers. The spine switch will be the BGP route reflector. I already configured the underlay network with OSPF, PIM, and the spine switch as the RP.

Configurations

Want to take a look for yourself? Here, you will find the startup configuration of each device.

SPINE1

hostname SPINE1

feature ospf
feature pim

ip pim rp-address 1.1.1.1 group-list 224.0.0.0/4
ip pim ssm range 232.0.0.0/8

interface Ethernet1/1
  no switchport
  mac-address 0050.c253.1001
  ip address 192.168.12.1/24
  ip ospf network point-to-point
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode
  no shutdown

interface Ethernet1/2
  no switchport
  mac-address 0050.c253.1002
  ip address 192.168.13.1/24
  ip ospf network point-to-point
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode
  no shutdown

interface loopback0
  ip address 1.1.1.1/32
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode
icam monitor scale

router ospf 1

LEAF1

hostname LEAF1

feature ospf
feature pim

ip pim rp-address 1.1.1.1 group-list 224.0.0.0/4
ip pim ssm range 232.0.0.0/8

interface Ethernet1/1
  no switchport
  mac-address 0050.c253.2001
  ip address 192.168.12.2/24
  ip ospf network point-to-point
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode
  no shutdown

interface Ethernet1/2
  switchport access vlan 10

interface loopback0
  ip address 2.2.2.2/32
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode

router ospf 1

LEAF2

hostname LEAF2

feature ospf
feature pim

ip pim rp-address 1.1.1.1 group-list 224.0.0.0/4
ip pim ssm range 232.0.0.0/8

interface Ethernet1/1
  no switchport
  mac-address 0050.c253.3001
  ip address 192.168.13.3/24
  ip ospf network point-to-point
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode
  no shutdown

interface Ethernet1/2
  switchport access vlan 10

interface loopback0
  ip address 3.3.3.3/32
  ip router ospf 1 area 0.0.0.0
  ip pim sparse-mode

router ospf 1





MP-BGP

Let’s start with the MP-BGP EVPN configuration. There are two global items we need to configure:

SPINE1, LEAF1 & LEAF2
(config)# feature bgp
(config)# nv overlay evpn

This is why we need to enable these two commands:

  • You need feature bgp so you can use the router bgp command.
  • You need nx overlay evpn otherwise, you can’t configure address-family l2vpn evpn under router bgp.

Now, we can configure MP-BGPN. I’ll start with the spine switch:

SPINE1(config)# router bgp 123
SPINE1(config-router)# neighbor 2.2.2.2
SPINE1(config-router-neighbor)# remote-as 123
SPINE1(config-router-neighbor)# update-source loopback 0
SPINE1(config-router-neighbor)# address-family l2vpn evpn
SPINE1(config-router-neighbor-af)# send-community 
SPINE1(config-router-neighbor-af)# send-community extended
SPINE1(config-router-neighbor-af)# route-reflector-client

SPINE1(config-router)# neighbor 3.3.3.3
SPINE1(config-router-neighbor)# remote-as 123
SPINE1(config-router-neighbor)# update-source loopback 0
SPINE1(config-router-neighbor)# address-family l2vpn evpn
SPINE1(config-router-neighbor-af)# send-community 
SPINE1(config-router-neighbor-af)# send-community extended 
SPINE1(config-router-neighbor-af)# route-reflector-client

This turns SPINE1 into a route reflector for the L2VPN EVPN address family. We also need extended communities enabled.

We’ll configure the leaf switches as route reflector clients:

LEAF1 & LEAF2
(config)# router bgp 123 
(config-router)# neighbor 1.1.1.1
(config-router-neighbor)# remote-as 123
(config-router-neighbor)# update-source loopback 0
(config-router-neighbor)# address-family l2vpn evpn
(config-router-neighbor-af)# send-community 
(config-router-neighbor-af)# send-community extended

that’s all for bgp…

VNI and NVE

Let’s configure the VNI and NVE interface. We’ll do this on both leaf switches:

LEAF1 & LEAF2

(config)#feature vn-segment-vlan-based
(config)#feature nv overlay

(config)# vlan 10
(config-vlan)# vn-segment 10010
(config-vlan)# exit

(config)# interface nve1
(config-if-nve)# no shutdown
(config-if-nve)# host-reachability protocol bgp
(config-if-nve)# source-interface loopback 0
(config-if-nve)# member vni 10010
(config-if-nve-vni)# mcast-group 239.1.1.1

This configuration is exactly the same as we have seen previously, with one exception. We now use the host-reachability protocol bgp command. This tells the leaf switches to learn MAC addresses in the control plane (using MP-BGP) and not on the data plane.

This completes our configuration.

Verification

Let’s verify our work. The NVE interface is up:

LEAF1# show nve vni
Codes: CP - Control Plane        DP - Data Plane          
       UC - Unconfigured         SA - Suppress ARP        
       SU - Suppress Unknown Unicast 
       Xconn - Crossconnect      
       MS-IR - Multisite Ingress Replication
 
Interface VNI      Multicast-group   State Mode Type [BD/VRF]      Flags
--------- -------- ----------------- ----- ---- ------------------ -----
nve1      10010    239.1.1.1         Up    CP   L2 [10]

Note that the mode now shows up as Control Plane (CP). Let’s send some traffic between the hosts:

We're Sorry, Full Content Access is for Members Only...

If you like to keep on reading, Become a Member Now! Here is why:

  • Learn any CCNA, CCNP and CCIE R&S Topic. Explained As Simple As Possible.
  • Try for Just $1. The Best Dollar You’ve Ever Spent on Your Cisco Career!
  • Full Access to our 799 Lessons. More Lessons Added Every Week!
  • Content created by Rene Molenaar (CCIE #41726)

567 Sign Ups in the last 30 days

satisfaction-guaranteed
100% Satisfaction Guaranteed!
You may cancel your monthly membership at any time.
No Questions Asked!

Forum Replies

  1. What if the spine switch had a different AS number?

  2. Hello @dimasgb ,

    There are some pros and cons of using eBGP or iBGP on the fabric. I’ll create an entire lesson for this with some other options for the underlay network. I’ll add a link here when it’s finished.

  3. Hello Amit

    Indeed MP-BGP EVPN resolves the issue of the flood and learn mechanism that multicast uses, enabling a more efficient communication using VXLAN. However, MP-BGP EVPN functions only for unicast traffic. Many services and devices will still be using BUM traffic (Broadcast, Unknown Unicast, Multicast), and this kind of traffic cannot be dealt with just using MP-BGP EVPN. There

    ... Continue reading in our forum

  4. Hello Amit

    Yes that’s correct, MP-BGP EVPN alone doesn’t natively handle BUM traffic. That’s why the multicast underlay must be configured in addition. It works alongside MP-BGP EVPN to handle the BUM traffic.

    I hope this has been helpful!

    Laz

  5. Hello :slight_smile: There is a line I never gave attention to, and now that I’m trying to understand it, i can’t find the logic ^^
    it is : Flags: (0x000102) (high32 00000000) on xmit-list, is not in l2rib/evpn

    Especially the not in l2rib, for a locally generated route. A local route injected into evpn comes necessarily from L2RIB. There must be a logic, but i can’t see it

    thanks a lot in advance
    cheers
    Imen

6 more replies! Ask a question or join the discussion by visiting our Community Forum